Socrates Technical Analysis Prediction Magic

The difficulties in following Martin Armstrong's Socrates trading advice

Why and how can Martin Armstrong and affiliates claim (make it appear to the untrained eye) that the Socrates system of Technical Analysis always creates the correct trading signals, makes the correct prediction and is free of human bias and never wrong?

The answer is very simple. The system produces multiple conflicting signals to choose from. In hindsight, one of the correct ones is chosen as the winner and the incorrect ones are ignored. That's it! We always win.

To explore how we get there, let us first look at the problem of forecasting from the rear vision mirror perspective - in hindsight.

Let's assume for argument's sake a scenario where we use a smart trading system that produces three possible signals from a single output: Buy, do nothing or sell.

We use that system and look in hindsight at a market situation, say a stock rally. We wanted to profit from it, but our system failed and advised to sell, so we sold and traded with a loss. Naturally, we review the situation and ask ourselves:

What was the signal that we missed that would have told us to trade in the right direction, the opposite direction?

Our system is a system that has only a single output, so there is no way we could have traded differently. No such signal!

Martin Armstrong would have had this problem when designing Socrates at the very beginning.

To solve this problem, Martin Armstrong would have extended his system to have multiple outputs to choose from. These outputs typically produce different results in any given market situation. Let us see how this is implemented in practice by inspecting the signals from a Socrates report in a single time frame, say daily, on the same day (quotes from actual reports):

• Our Daily level momentum is bullish
• The trend indicator is neutral
• Our long-term trend is bearish
• The cyclical strength indicator is bullish
• The market is trading above one indicating pivot implying that this market is in a positive position
• With respect to time, there is a prospect of a decline moving into (the following day)
• The overall tone on the Daily level is negative since the last Daily Bearish Reversal being elected ...
• The broader tone of the market is still positive on this trading level.
• The Oscillators are in a mixed position.
• Immediately, our (Energy) model continues to rally suggesting that a strong rally is likely. 
• We have elected a Bearish / Bullish reversal

Additionally, we may get:

•  Fake elected Reversals count
•  Superposition event canceling a recently elected reversal (Refer to Revision Signals)

These can all appear on the same day.

Then there are signals that must be considered but they are not in the reports:

•  The NON-election of a nearby Reversal (the most mysterious "signal" one could imagine)
•  Reversal One Percent Rule
•  Reversal gap

We are counting 16 signals in a single time frame. When we multiply this by 3 for daily, weekly and monthly time frames to be conservative, we get 48 signals to choose from, each of them may have different directions.

The Socrates system does not provide a computerized solution for consolidating these signals into a meaningful response, so it is a matter of human interpretation to find the correct conclusion.

Now let us look how this expansion is exploited to always produce the correct result in hindsight, by means of an example:

Wheat Futures Example

Let's try again to trade it, this time monthly reversals. Date elected: 2019-04-30, traded for one month until 2019-05-31
GDAX monthly bullish 5% loss
GDX monthly bearish 3.4% loss
SP 500 monthly bullish 6.6% loss
TLT monthly bearish 6.6% loss
/ZW Wheat futures monthly bearish 17.4% loss

On Friday, 2019-04-30 /ZW Wheat Futures definitely elected a Monthly Bearish Reversal, that, if had it been traded for the month, would have produced a loss of 17.4%. Martin Armstrong's response to this dilemma is to gaslight (blame) the user as follows:

Martin Armstrong's Response:

I saw your incredibly uninformed comment on wheat futures if you had the trader service you would know that MAY was a turning point and there was a MAJOR monthly bearish reversal just below which was not elected the only charlatan here is you.

In clear text, this means that in hindsight, the correct conclusion would have been to ignore the ELECTED Monthly Bearish Reversal, and instead to pay attention to the NOT ELECTED next reversal, i.e. a missing signal.

The undisputed fact is that at this time, the implied rule, the "Elected Reversal Invalidation Rule" as I call it, which forces the invalidation or negation of an elected Reversal by another non-elected Reversal, does not exist in any of the Socrates reference documentation The Reversal System.

In addition, the user needs to consider the turning point, the timing information from the forecast arrays. There is no signal that the system puts out that would indicate this choice. The user would have had to carry out research to come to this extremely far-fetched conclusion. A system of maximum effort for negative returns.

The answer to the question we asked in the beginning is fairly obvious from this example. The conclusion for the correct trading decision can only be made in hindsight by cherry-picking some signals that best fit the situation in the past. That is because there are no rules for how to pick the best mix of these signals for a trade. These rules are made up by Martin Armstrong and his staff in hindsight on the fly to match any situation as required.

Therefore, the Socrates system cannot be used to get accurate and verifiable trading decisions in real time.

See also:

The Reversal System - Engineering Background